Structural analysis of the time data series

Consideration was given to different methods of structural analysis of the data series which are intended for expeditious determination of the “special points” – the instants of changes in trends. The data about different physical units such as technical, market, and other objects feature great dynamic variety in terms of their trends, local structural characteristics, and sliding statistical characteristics. Expeditious recognition of the instants of changes in the realization characteristics enables one to make timely decisions about normal operation of the objects under consideration. A multivariant approach to structural analysis of data realizations was proposed to enhance the reliability of determining the instants of occurrence of special points. Particular examples of multivariant structural analysis were presented.

Presented by the member of Editorial Board: V. V. Kul'ba